Jason Chen on Coding for a Deeper Read of Disclosures
Jason Chen on Coding for a Deeper Read of Disclosures Heading link
Accounting information isn’t all numbers. Managers reveal a lot about their firms in the language they use; this is where Jason Chen’s research comes in. “Anyone who’s picked up an annual report knows that accounting information isn’t strictly numerical. A lot of this information is in the form of managers talking about the company. They talk about what they’ve done and what they plan to do. This information is important, and isn’t captured quantitatively.”
Chen, who grew up in the Chicago area and obtained his MSA from UIC in 2009, says he was drawn back to UIC by the quality of support for students and faculty research. Much of the work he’ll be doing at UIC involves building computer programs that extract information about everything from accruals to how analysts obtain information.
“Transcripts can be read for the questions that analysts ask in a conference call, which tell us something about the analyst’s beliefs, and these can give clues to changes in stock price following a firm’s disclosure.”Chen approached qualitative analysis from his background in computer science. “I still read a lot of text myself – 10Ks, MD&As, and transcripts of calls, but once I have a strong intuitive understanding of how the text is working, I start coding.”
Ultimately, Chen’s programs process huge volumes of documents in order to pull out key qualitative disclosures. Despite the technical detail required in this work, he doesn’t let students miss the forest for the trees. “One of my goals in teaching is to encourage critical thinking. Students can get caught up in technical details; these details are important, but understanding the arguments that underpin them is what broadens a student’s accounting knowledge.”